For a molecule with 2 chiral centres, there are four possible combinations, (R,R), (R,S), (S,R) and (S,S). For example, 3-bromo-2-butanol has four stereoisomers: (2R,3R)-3-bromo-2-butanol, (2R,3S)-3-bromo-2-butanol, (2S,3R)-3-bromo-2-butanol and (2S,3S)-3-bromo-2-butanol.
Of these combinations, the following pairs are diastereomers:
S,R and S,S
S,R and R,R
R,S and S,S
R,S and R,R
I notice that only one chiral centre has been flipped to form these pairs.
The following pairs are either enantiomers or meso compounds (e.g. for (2R,3S)-tartaric acid):
S,R and R,S
S,S and R,R
I notice that both chiral centres have been flipped to form these pairs.
Could we generalise that flipping an even number of chiral centres always gives either an enantiomer or the same molecule if it's a meso compound, and flipping an odd number of chiral centres always gives a diastereomer?